A role for the ecological study in the developing world.

Retrospective case-control or prospective (follow-up) studies are important epidemiological tools and have provided useful information on exposure disease associations. Prospective studies would be the ideal option, but many countries (particularly in the developing world) do not have the necessary infrastructure to follow people up. Both retrospective and prospective studies are, however, sometimes conducted without due regard for their own limitations. These limitations are exacerbated when measures of exposure or disease are based on a single measurement and where the population under study is homogeneous with regard to exposure. The former is responsible for regression dilution bias and the latter for a lack of contrasts between exposure groups. Both factors would attenuate any relationship between exposure and disease. Ecological studies in epidemiology are weaker in design than case-control or prospective studies, but in some circumstances an ecological approach, which looks at the prevalence of an exposure or disorder in a number of areas of varying disease rates, may offer some advantages.